Rules Engine
The rules engine is the core that checks your shot data and symptoms against sport-specific target windows to find the best-fitting pattern.
The rules engine is the configuration layer that gives the heuristic engine its sport-specific knowledge. It is built as a set of explicit, version-controlled rules. Each rule defines a condition, such as a metric value outside a target window, a symptom pattern, or a combination of observable cues. It also defines the fault that condition indicates, the severity of that fault, and the recommended initial response. When the engine receives your data, it evaluates your values against each rule. It returns the highest-scoring match as the primary finding. Any patterns that appear alongside it are noted as secondary.
Because the rules are explicit and version-controlled, the results are repeatable and inspectable. You can see each value beside the target range it was judged against. This traceability is a design requirement, not an afterthought. SwingVantage is committed to showing the reasoning behind every finding, and the rules engine is what makes that possible. A rule that is wrong or outdated can be corrected without retraining a model. The correction applies to later analyses.
Score and grade outputs use a fixed pro standard. Stated skill level may change the coaching language or practice prescription, but it does not curve the result. The report shows the target used and leaves unsupported metrics unscored. It also lets a genuinely poor measured result fall to zero instead of applying a positive floor.
Tips by skill level
If a finding surprises you, look at your value and the fixed pro target shown in the report. Your skill level personalizes the coaching plan, not the score.
If you are reliably at the edge of a threshold — right on the boundary of a flagged window — pay attention to whether the finding triggers on some sessions and not others. This variability is often more informative than a consistent flag. It suggests the fault is present intermittently. This can point to a timing or consistency issue rather than a fundamental technique problem.
Example
A driver session's average spin sits well above its target window, so the rules engine flags it and shows your average beside the target range.
How it shows up on video
The rules engine output is visible in the diagnosis section of your report. Each finding states your values, and its evidence lists each metric beside the target range it was judged against. That makes the rule-based reasoning transparent.
Common mistakes
- Conflating the rules engine with generic advice — the rules in the engine are built on sport-specific coaching knowledge and benchmarks, not general tips.
- Expecting the rules engine to disagree with itself across sessions — by design, the same inputs always produce the same outputs. If findings vary across sessions, the input data (video quality, swing conditions) varied.
- Not updating practice when the rules engine flags the same fault across multiple sessions. A recurring fault finding is the engine signaling that the primary issue has not resolved.
- Assuming a finding from the rules engine is less credible than a finding from an AI model — rule-based findings are explicitly grounded in coaching knowledge and are fully traceable. AI-augmented findings add narrative depth but are built on top of the rule-based foundation.
In SwingVantage Motion Lab
When you view a finding in SwingVantage, the rules engine is what produced it. The fault label, priority, and target shown in each metric's detail all come from the rules engine's evaluation of your data. When an update changes how a read is measured, Motion Lab says so and does not compare clips analyzed before and after it.
Frequently asked questions
Where do the rules engine's benchmarks come from?
Benchmarks come from published sports science research and coaching knowledge, not proprietary or fabricated databases. Where research is thin, a Motion Lab target is labeled a Starter estimate rather than presented as a validated norm. SwingVantage does not invent benchmarks; it organizes and applies existing expert knowledge.
Can the rules engine handle faults that do not fit a standard pattern?
The rules engine handles the fault patterns it was built with — common, well-documented issues that appear across large populations of athletes. An unusual or compound fault that fits no standard pattern gets a weaker read. The report says so plainly, gives general guidance, and suggests what would firm it up, such as a video or a retest.
Related terms
- Deterministic IntelligenceDeterministic intelligence is analysis that follows explicit, written rules. The same inputs always produce the same answer, and you can trace exactly why.
- BenchmarkA benchmark is a reference value that defines acceptable or optimal performance for a metric; SwingVantage score outputs use one fixed pro grading standard.
- Heuristic EngineThe heuristic engine is the deterministic, rules-based part of SwingVantage that runs first on every analysis — no external AI call required.
See it in a real analysis
Try SwingVantage on one of your own swings — free to start.